arXiv subjects

Wei Li

Publications and source records attributed to Wei Li.

At least 829 records · Page 46Linked to original sources

arXiv2022

Misaligned orientations of 4f optical neural network for image classification accuracy on various datasets

In recent years, the optical 4f system has drawn much attention in building high-speed and ultra-low-power optical neural networks (ONNs). Most optical systems suffer from the misalignment of the optical devices during installment. The performance of ONN based on the optical 4f system (4f-ONN) is considered sensitive to the misalignment in the optical path introduced. In order to comprehensively investigate the influence caused by the misalignment, we proposed a method for estimating the performance of a 4f-ONN in response to various misalignment in the context of the image classification task.The misalignment in numerical simulation is estimated by manipulating the optical intensity distributions in the fourth focus plane in the 4f system. Followed by a series of physical experiments to validate the simulation results. Using our method to test the impact of misalignment of 4f system on the classification accuracy of two popular image classification datasets, MNIST and Quickdraw16. On both datasets, we found that the performances of 4f-ONN generally degraded dramatically as the positioning error increased. Different positioning error tolerance in the misalignment orientations was observed over the two datasets. Classification performance could be preserved by positioning errors up to 200 microns in a specific direction.

physics.optics↗

arXiv2022

SoccerNet 2022 Challenges Results

The SoccerNet 2022 challenges were the second annual video understanding challenges organized by the SoccerNet team. In 2022, the challenges were composed of 6 vision-based tasks: (1) action spotting, focusing on retrieving action timestamps in long untrimmed videos, (2) replay grounding, focusing on retrieving the live moment of an action shown in a replay, (3) pitch localization, focusing on detecting line and goal part elements, (4) camera calibration, dedicated to retrieving the intrinsic and extrinsic camera parameters, (5) player re-identification, focusing on retrieving the same players across multiple views, and (6) multiple object tracking, focusing on tracking players and the ball through unedited video streams. Compared to last year's challenges, tasks (1-2) had their evaluation metrics redefined to consider tighter temporal accuracies, and tasks (3-6) were novel, including their underlying data and annotations. More information on the tasks, challenges and leaderboards are available on https://www.soccer-net.org. Baselines and development kits are available on https://github.com/SoccerNet.

cs.CV↗

arXiv2022

Bidirectional flow of two-dimensional dusty plasma under asymmetric periodic substrates driven by unbiased external excitations

Collective transport properties of a one-dimensional asymmetric periodic substrate (1DAPS) modulated two-dimensional dusty plasma (2DDP) driven by an unbiased sinusoidal excitation force are investigated using Langevin dynamical simulations. It is discovered that, by changing the amplitude and frequency of the unbiased sinusoidal external excitations, as well as the depth of 1DPAS, both the direction and speed of the persistent particle flow can be adjusted, i.e., both the flow rectification and its reversal of the ratchet effect of the steady drift motion for particles are achieved using various excitations. For the studied 2DDP under the 1DAPS, when the amplitude of the excitation increases from zero, the magnitude of the overall drift velocity increases from zero to its first maximum in the easy direction of the 1DAPS, next decreases gradually back to zero, and then increases from zero to its second maximum in the hard direction of 1DAPS before finally gradually decaying. It is found that, as the frequency of the excitation and the depth of 1DAPS change, the maximum overall drift velocity also varies, and the corresponding magnitude of the excitation varies simultaneously. The observed ratchet effect in both the easy and hard directions of 1DAPS for 2DDP is attributed to the combination of the spatial symmetry breaking of 1DAPS and the inertial effects of particles, which is further confirmed by the three different presented diagnostics.

physics.plasm-ph↗

arXiv2022

Multi-Grained Angle Representation for Remote Sensing Object Detection

Arbitrary-oriented object detection (AOOD) plays a significant role for image understanding in remote sensing scenarios. The existing AOOD methods face the challenges of ambiguity and high costs in angle representation. To this end, a multi-grained angle representation (MGAR) method, consisting of coarse-grained angle classification (CAC) and fine-grained angle regression (FAR), is proposed. Specifically, the designed CAC avoids the ambiguity of angle prediction by discrete angular encoding (DAE) and reduces complexity by coarsening the granularity of DAE. Based on CAC, FAR is developed to refine the angle prediction with much lower costs than narrowing the granularity of DAE. Furthermore, an Intersection over Union (IoU) aware FAR-Loss (IFL) is designed to improve accuracy of angle prediction using an adaptive re-weighting mechanism guided by IoU. Extensive experiments are performed on several public remote sensing datasets, which demonstrate the effectiveness of the proposed MGAR. Moreover, experiments on embedded devices demonstrate that the proposed MGAR is also friendly for lightweight deployments.

cs.CV↗

arXiv2022

Obj2Seq: Formatting Objects as Sequences with Class Prompt for Visual Tasks

Visual tasks vary a lot in their output formats and concerned contents, therefore it is hard to process them with an identical structure. One main obstacle lies in the high-dimensional outputs in object-level visual tasks. In this paper, we propose an object-centric vision framework, Obj2Seq. Obj2Seq takes objects as basic units, and regards most object-level visual tasks as sequence generation problems of objects. Therefore, these visual tasks can be decoupled into two steps. First recognize objects of given categories, and then generate a sequence for each of these objects. The definition of the output sequences varies for different tasks, and the model is supervised by matching these sequences with ground-truth targets. Obj2Seq is able to flexibly determine input categories to satisfy customized requirements, and be easily extended to different visual tasks. When experimenting on MS COCO, Obj2Seq achieves 45.7% AP on object detection, 89.0% AP on multi-label classification and 65.0% AP on human pose estimation. These results demonstrate its potential to be generally applied to different visual tasks. Code has been made available at: https://github.com/CASIA-IVA-Lab/Obj2Seq.

cs.CV↗

arXiv2022

Quantum Spin Liquid with Emergent Chiral Order in the Triangular-lattice Hubbard Model

The interplay between spin frustration and charge fluctuation gives rise to an exotic quantum state in the intermediate-interaction regime of the half-filled triangular-lattice Hubbard model, while the nature of the state is under debate. Using the density matrix renormalization group with SU(2)$_{\rm{spin}} \otimes $U(1)$_{\rm{charge}}$ symmetries implemented, we study the triangular-lattice Hubbard model defined on the long cylinder geometry up to circumference $W=6$. A gapped quantum spin liquid, with on-site interaction $9 \lesssim U / t \lesssim 10.75$, is identified between the metallic and the antiferromagnetic Mott insulating phases. In particular, we find that this spin liquid develops a robust long-range spin scalar-chiral correlation as the system length $L$ increases, which unambiguously unveils the spontaneous time-reversal symmetry breaking. In addition, the degeneracy of the entanglement spectrum supports symmetry fractionalization and spinon edge modes in the obtained ground state. The possible origin of chiral order in this intermediate spin liquid and its relation to the rotonlike excitations have also been discussed.

cond-mat.str-el↗

arXiv2022

Compositionally Complex Perovskite Oxides for Solar Thermochemical Water Splitting

Solar thermochemical hydrogen generation (STCH) is a promising approach for eco-friendly H2 production, but conventional STCH redox compounds often suffer from thermodynamic and kinetic limitations with limited tunability. Expanding from the nascent high-entropy ceramics field, this study explores a new class of compositionally complex perovskite oxides (La0.8Sr0.2)(Mn(1-x)/3Fe(1-x)/3CoxAl(1-x)/3)O3 for STCH. In situ X-ray diffraction demonstrates the phase stability during redox cycling and in situ X-ray photoelectron spectroscopy shows preferential redox of Co. The extent of reduction increases, but the intrinsic kinetics decreases, with increased Co content. Consequently, (La0.8Sr0.2)(Mn0.2Fe0.2Co0.4Al0.2)O3-δ achieves an optimal balance between the thermodynamics and kinetics properties. The combination of a moderate enthalpy of reduction, high entropy of reduction, and preferable surface oxygen exchange kinetics enables a maximum H2 yield of 395 +- 11 μmol g-1 in a short 1-hour redox duration. Entropy stabilization expectedly contributes to the structure stability during redox without phase transformation, which enables an exceptional STCH stability for >50 cycles under harsh interrupted conditions. The underlying redox mechanism is further elucidated by the density functional theory based parallel Monte Carlo computation, which represents a new computation paradigm first established here. This study suggests a new class of non-equimolar compositionally complex ceramics for STCH and chemical looping.

cond-mat.mtrl-sci↗

arXiv2022

Playing Technique Detection by Fusing Note Onset Information in Guzheng Performance

The Guzheng is a kind of traditional Chinese instruments with diverse playing techniques. Instrument playing techniques (IPT) play an important role in musical performance. However, most of the existing works for IPT detection show low efficiency for variable-length audio and provide no assurance in the generalization as they rely on a single sound bank for training and testing. In this study, we propose an end-to-end Guzheng playing technique detection system using Fully Convolutional Networks that can be applied to variable-length audio. Because each Guzheng playing technique is applied to a note, a dedicated onset detector is trained to divide an audio into several notes and its predictions are fused with frame-wise IPT predictions. During fusion, we add the IPT predictions frame by frame inside each note and get the IPT with the highest probability within each note as the final output of that note. We create a new dataset named GZ_IsoTech from multiple sound banks and real-world recordings for Guzheng performance analysis. Our approach achieves 87.97% in frame-level accuracy and 80.76% in note-level F1-score, outperforming existing works by a large margin, which indicates the effectiveness of our proposed method in IPT detection.

cs.SD↗

arXiv2022

Learning from Future: A Novel Self-Training Framework for Semantic Segmentation

Self-training has shown great potential in semi-supervised learning. Its core idea is to use the model learned on labeled data to generate pseudo-labels for unlabeled samples, and in turn teach itself. To obtain valid supervision, active attempts typically employ a momentum teacher for pseudo-label prediction yet observe the confirmation bias issue, where the incorrect predictions may provide wrong supervision signals and get accumulated in the training process. The primary cause of such a drawback is that the prevailing self-training framework acts as guiding the current state with previous knowledge, because the teacher is updated with the past student only. To alleviate this problem, we propose a novel self-training strategy, which allows the model to learn from the future. Concretely, at each training step, we first virtually optimize the student (i.e., caching the gradients without applying them to the model weights), then update the teacher with the virtual future student, and finally ask the teacher to produce pseudo-labels for the current student as the guidance. In this way, we manage to improve the quality of pseudo-labels and thus boost the performance. We also develop two variants of our future-self-training (FST) framework through peeping at the future both deeply (FST-D) and widely (FST-W). Taking the tasks of unsupervised domain adaptive semantic segmentation and semi-supervised semantic segmentation as the instances, we experimentally demonstrate the effectiveness and superiority of our approach under a wide range of settings. Code will be made publicly available.

cs.CV↗

arXiv2022

Evolutionary Action Selection for Gradient-based Policy Learning

Evolutionary Algorithms (EAs) and Deep Reinforcement Learning (DRL) have recently been integrated to take the advantage of the both methods for better exploration and exploitation.The evolutionary part in these hybrid methods maintains a population of policy networks.However, existing methods focus on optimizing the parameters of policy network, which is usually high-dimensional and tricky for EA.In this paper, we shift the target of evolution from high-dimensional parameter space to low-dimensional action space.We propose Evolutionary Action Selection-Twin Delayed Deep Deterministic Policy Gradient (EAS-TD3), a novel hybrid method of EA and DRL.In EAS, we focus on optimizing the action chosen by the policy network and attempt to obtain high-quality actions to promote policy learning through an evolutionary algorithm. We conduct several experiments on challenging continuous control tasks.The result shows that EAS-TD3 shows superior performance over other state-of-art methods.

cs.NE↗

arXiv2022

LO-Det: Lightweight Oriented Object Detection in Remote Sensing Images

A few lightweight convolutional neural network (CNN) models have been recently designed for remote sensing object detection (RSOD). However, most of them simply replace vanilla convolutions with stacked separable convolutions, which may not be efficient due to a lot of precision losses and may not be able to detect oriented bounding boxes (OBB). Also, the existing OBB detection methods are difficult to constrain the shape of objects predicted by CNNs accurately. In this paper, we propose an effective lightweight oriented object detector (LO-Det). Specifically, a channel separation-aggregation (CSA) structure is designed to simplify the complexity of stacked separable convolutions, and a dynamic receptive field (DRF) mechanism is developed to maintain high accuracy by customizing the convolution kernel and its perception range dynamically when reducing the network complexity. The CSA-DRF component optimizes efficiency while maintaining high accuracy. Then, a diagonal support constraint head (DSC-Head) component is designed to detect OBBs and constrain their shapes more accurately and stably. Extensive experiments on public datasets demonstrate that the proposed LO-Det can run very fast even on embedded devices with the competitive accuracy of detecting oriented objects.

cs.CV↗

arXiv2022

Spin Supersolidity in Nearly Ideal Easy-axis Triangular Quantum Antiferromagnet Na$_2$BaCo(PO$_4$)$_2$

Prototypical models and their material incarnations are cornerstones to the understanding of quantum magnetism. Here we show theoretically that the recently synthesized magnetic compound Na$_2$BaCo(PO$_4$)$_2$ (NBCP) is a rare, nearly ideal material realization of the $S=1/2$ triangular-lattice antiferromagnet with significant easy-axis spin exchange anisotropy. By combining the automatic parameter searching and tensor-network simulations, we establish a microscopic model description of this material with realistic model parameters, which can not only fit well the experimental thermodynamic data but also reproduce the measured magnetization curves without further adjustment of parameters. According to the established model, the NBCP hosts a spin supersolid state that breaks both the lattice translation symmetry and the spin rotational symmetry. Such a state is a spin analogue of the long-sought supersolid state, thought to exist in solid Helium and optical lattice systems, and share similar traits. The NBCP therefore represents an ideal material-based platform to explore the physics of supersolidity as well as its quantum and thermal melting.

cond-mat.str-el↗

arXiv2022

Language-aware Domain Generalization Network for Cross-Scene Hyperspectral Image Classification

Text information including extensive prior knowledge about land cover classes has been ignored in hyperspectral image classification (HSI) tasks. It is necessary to explore the effectiveness of linguistic mode in assisting HSI classification. In addition, the large-scale pre-training image-text foundation models have demonstrated great performance in a variety of downstream applications, including zero-shot transfer. However, most domain generalization methods have never addressed mining linguistic modal knowledge to improve the generalization performance of model. To compensate for the inadequacies listed above, a Language-aware Domain Generalization Network (LDGnet) is proposed to learn cross-domain invariant representation from cross-domain shared prior knowledge. The proposed method only trains on the source domain (SD) and then transfers the model to the target domain (TD). The dual-stream architecture including image encoder and text encoder is used to extract visual and linguistic features, in which coarse-grained and fine-grained text representations are designed to extract two levels of linguistic features. Furthermore, linguistic features are used as cross-domain shared semantic space, and visual-linguistic alignment is completed by supervised contrastive learning in semantic space. Extensive experiments on three datasets demonstrate the superiority of the proposed method when compared with state-of-the-art techniques.

cs.CV↗

arXiv2022

Discrete Boltzmann multi-scale modeling of non-equilibrium multiphase flows

The aim of this paper is twofold: the first is to formulate and validate a multi-scale discrete Boltzmann method (DBM) based on density functional kinetic theory for thermal multiphase flow systems, ranging from continuum to transition flow regime; the second is to present some new insights into the thermo-hydrodynamic non-equilibrium (THNE) effects in the phase separation process. Methodologically, DBM includes three main pillars: (i) the determination of the fewest kinetic moment relations, which are required by the description of significant THNE effects beyond the realm of continuum fluid mechanics, (ii) the construction of appropriate discrete equilibrium distribution function recovering all the desired kinetic moments, (iii) the detection, description, presentation and analysis of THNE based on the moments of the non-equilibrium distribution ($f-f^{(eq)}$). The incorporation of appropriate additional higher-order thermodynamic kinetic moments considerably extends the DBM's capability of handling larger values of the liquid-vapor density ratio, curbing spurious currents, and ensuring mass-momentum-energy conservation. Compared with the DBM with only first-order THNE (Gan et al. Soft Matter 11,5336), the model retrieves kinetic moments beyond the third-order super-Burnett level, and is accurate for weak, moderate, and strong THNE cases even when the local Knudsen number exceeds $1/3$. Physically, the ending point of the linear relation between THNE and the concerned physical parameter provides a distinct criterion to identify whether the system is near or far from equilibrium. Besides, the surface tension refrains the local THNE around the interface, but expands the THNE range and strengthens the THNE intensity away from the interface through interface smoothing and widening.

physics.flu-dyn↗

arXiv2022

Single-source Domain Expansion Network for Cross-Scene Hyperspectral Image Classification

Currently, cross-scene hyperspectral image (HSI) classification has drawn increasing attention. It is necessary to train a model only on source domain (SD) and directly transferring the model to target domain (TD), when TD needs to be processed in real time and cannot be reused for training. Based on the idea of domain generalization, a Single-source Domain Expansion Network (SDEnet) is developed to ensure the reliability and effectiveness of domain extension. The method uses generative adversarial learning to train in SD and test in TD. A generator including semantic encoder and morph encoder is designed to generate the extended domain (ED) based on encoder-randomization-decoder architecture, where spatial and spectral randomization are specifically used to generate variable spatial and spectral information, and the morphological knowledge is implicitly applied as domain invariant information during domain expansion. Furthermore, the supervised contrastive learning is employed in the discriminator to learn class-wise domain invariant representation, which drives intra-class samples of SD and ED. Meanwhile, adversarial training is designed to optimize the generator to drive intra-class samples of SD and ED to be separated. Extensive experiments on two public HSI datasets and one additional multispectral image (MSI) dataset demonstrate the superiority of the proposed method when compared with state-of-the-art techniques.

cs.CV↗

arXiv2022

Towards Accurate Binary Neural Networks via Modeling Contextual Dependencies

Existing Binary Neural Networks (BNNs) mainly operate on local convolutions with binarization function. However, such simple bit operations lack the ability of modeling contextual dependencies, which is critical for learning discriminative deep representations in vision models. In this work, we tackle this issue by presenting new designs of binary neural modules, which enables BNNs to learn effective contextual dependencies. First, we propose a binary multi-layer perceptron (MLP) block as an alternative to binary convolution blocks to directly model contextual dependencies. Both short-range and long-range feature dependencies are modeled by binary MLPs, where the former provides local inductive bias and the latter breaks limited receptive field in binary convolutions. Second, to improve the robustness of binary models with contextual dependencies, we compute the contextual dynamic embeddings to determine the binarization thresholds in general binary convolutional blocks. Armed with our binary MLP blocks and improved binary convolution, we build the BNNs with explicit Contextual Dependency modeling, termed as BCDNet. On the standard ImageNet-1K classification benchmark, the BCDNet achieves 72.3% Top-1 accuracy and outperforms leading binary methods by a large margin. In particular, the proposed BCDNet exceeds the state-of-the-art ReActNet-A by 2.9% Top-1 accuracy with similar operations. Our code is available at https://github.com/Sense-GVT/BCDN

cs.CV↗

arXiv2022

HPPNet: Modeling the Harmonic Structure and Pitch Invariance in Piano Transcription

While neural network models are making significant progress in piano transcription, they are becoming more resource-consuming due to requiring larger model size and more computing power. In this paper, we attempt to apply more prior about piano to reduce model size and improve the transcription performance. The sound of a piano note contains various overtones, and the pitch of a key does not change over time. To make full use of such latent information, we propose HPPNet that using the Harmonic Dilated Convolution to capture the harmonic structures and the Frequency Grouped Recurrent Neural Network to model the pitch-invariance over time. Experimental results on the MAESTRO dataset show that our piano transcription system achieves state-of-the-art performance both in frame and note scores (frame F1 93.15%, note F1 97.18%). Moreover, the model size is much smaller than the previous state-of-the-art deep learning models.

cs.SD↗

arXiv2022

The partial waves of $B^{*}_{2}(5747)$ and their contributions in strong decays

By adopting the relativistic Bethe-Salpeter method, the OZI allowed strong decays of the $2^+$ state $B^{*}_{2}(5747)^{0}$ are studied, emphasis is paid to the relativistic corrections. We first study the partial waves in the wave functions used, find that there are $P$, $D$ and $F$ partial waves in $B^{*}_{2}(5747)^{0}$ meson, and the ratios $P:D:F=1:0.421:0.051$. We also find $S:P:D=1:0.354:0.046$ for $B^*$, and $S:P=1:0.343$ for $B$ meson. The large components of the $D$ wave in $B^{*}_{2}(5747)^{0}$ and $P$ wave in $B^{(*)}$ means that large relativistic effects existing in these states. Second, we calculate the strong decays, the obtained total decay width $Γ(B^{*}_{2}(5747)^{0})=25.9$ MeV and the branching fraction $Γ(B^{*}_{2}(5747)^{0} \to B^{*}π)$ / $Γ(B^{*}_{2}(5747)^{0} \to Bπ)=0.96$ are consistent well with experimental data. Third, we study the contributions of different partial waves in the initial and final wave functions, find that the relativistic effects are about $15\%$ and $11\%$ for $B^{*}_{2}(5747)^{0} \to Bπ$ and $B^{*}_{2}(5747)^{0} \to B^{*}π$, respectively, which are much smaller than our expected, showing that the relativistic corrections are cancelled to each other in these decays.

hep-ph↗